1. Introduction
Survey of historical tropical cyclone (TC) tracks over the western North Pacific (WNP) indicates that west-moving and northwest-moving TCs occasionally experienced sudden northward track changes, which typically consist of rapid slowing of the westward movement and a substantial northward acceleration. Carr and Elsberry (1995) argued that sudden northward track changes over the WNP are usually accompanied by the coalescence of a TC with a large-scale monsoon gyre. The latter is a specific pattern of the evolution of low-level monsoon circulation and can be identified as a low-frequency, nearly circular cyclonic vortex with a diameter of about 2500 km (Lander 1994; Carr and Elsberry 1995). As a result of Rossby wave energy dispersion associated with the monsoon gyre, Carr and Elsberry (1995) numerically demonstrated that strong ridging (negative vorticity tendency) occurs to the southeast of a barotropic monsoon gyre. As the center of a TC is nearly collocated with the center of the monsoon gyre, the southwesterly winds between the monsoon gyre and the resulting anticyclonic circulation are substantially enhanced, leading to sudden northward turning in the TC track.
The TCs in the WNP are usually accompanied by a cyclonic circulation on the time scales ranging from intraseasonal oscillations to synoptic disturbances. For example, Ko and Hsu (2006) found that recurving TCs were often embedded in the cyclonic circulation of a submonthly wave pattern, which propagated north-northwestward over the WNP. With a focus on roles of slowly varying low-frequency monsoon circulations, Wu et al. (2011a) recently conducted an observational analysis of four typhoons that took a generally northwestward track prior to sharply turning northeastward in the vicinity of the East China Sea. They found that sudden track changes occurred near the center of a monsoon gyre or at the bifurcation point of the low-level steering flows on the Madden–Julian oscillation (MJO) time scale, but were all associated with the coalescence with a large-scale monsoon gyre on the quasi-biweekly oscillation (QBW) time scale. In agreement with Carr and Elsberry (1995), Wu et al. (2011a) suggest that the enhanced synoptic-scale southwesterly flows shifted the typhoons northward and placed them in a northeastward orbit. Wu et al. (2011b) and Liang et al. (2011) investigated the causes of the sudden northward track changes and the unusually long residence time of Typhoon Morakot (2009) in the vicinity of Taiwan. The typhoon generally moved westward prior to its landfall on Taiwan, and underwent a coalescence process first with a cyclonic gyre on the QBW time scale and then with a cyclonic gyre on the MJO time scale. Their observational and numerical studies confirm that the coalescences enhanced the synoptic-scale southwesterly winds of Morakot and thus decreased its westward movement and turned the track northward, leading to an unusually long residence time in the vicinity of Taiwan.
Given these studies that suggest sudden poleward track changes result from interaction between TCs and low-frequency large-scale monsoon gyres, a question arises as to whether low-frequency monsoon gyres and the enhanced synoptic-scale southwesterly winds are a typical flow pattern associated with all of the sudden northward track changes over the WNP. The main objective of this study is to address this issue. First we define the sudden north-turning track changes over the WNP and identify all of those cases that occurred during the period of May–November 2000–10. As an opposite track-turning type, we also define the west-turning track change although the magnitude of the direction change is relatively small. Together with the description of the datasets used in this study, the identification of sudden north-turning track changes as well as west-turning track changes is described in section 2, followed by an analysis of the operational forecast errors of the track change cases in the National Meteorological Center of the China Meteorological Administration (NMC/CMA) in section 3. In sections 4 and 5, we show that the identified TC track changes are closely associated with interactions of multitime scale flows. The roles of the interaction between TCs and low-frequency circulation in the TC track changes is discussed based on an analysis of vorticity budget in section 6, followed by a brief summary in section 7.
2. Data and identification of sudden TC track changes
In this study we only discuss TCs in the WNP basin with maximum sustained wind exceeding 17.2 m s−1. The track changes are identified from the Japan Meteorological Agency (JMA) best-track dataset, which includes the TC center position (latitude and longitude), the maximum sustained wind speed, and the minimum sea level pressure. The wind fields associated with sudden track changes are based on the National Centers for Environmental Prediction (NCEP) Final (FNL) operational global analysis data on 1.0° × 1.0° grids at every 6 h. Note that although the inner circulation of TCs is not reliable in the FNL dataset because of its relatively coarse horizontal resolution, previous studies have demonstrated that TC motion is relevant mainly to the outer asymmetric structure (Holland 1983; Chan and Williams 1987), which is susceptible to the environmental influences (Merrill 1984; Holland and Merrill 1984). Our calculation of steering flows in section 4 also suggests that the FNL wind data are suitable for this study.
Previous studies show that three prominent categories of atmospheric variations associated with TC activity are dominant in the tropical Pacific: intraseasonal oscillation (ISO) on the time scale of 40–50 or 30–60 days (hereafter the MJO time scale; Madden and Julian 1971, 1972, 1994), intraseasonal oscillation on the time scale of 10–20 days (hereafter the QBW time scale; Murakami and Frydrych 1974; Murakami 1976; Krishnamurti and Ardanuy 1980; Chen and Chen 1993; Kiladis and Wheeler 1995), and the synoptic-scale disturbances with periods between 3 and 10 days (Lau and Lau 1990; Chang et al. 1996). The ISOs are closely associated with the activity of Asia summer monsoon (Yasunari 1981; Chen et al. 2000; Chan et al. 2002; Goswami et al. 2003). To examine the influences of these multitime-scale environmental flows on sudden TC track changes, we use Lanczos filters in time at each grid point (Duchon 1979). A low-pass filter with a 10-day cutoff period is used to isolate the ISO time scale and the background state. The synoptic-scale circulation is then obtained as the difference between the unfiltered field and the filtered with the 10-day low-pass filter, including TC circulation in the FNL data. In addition, the MJO time-scale flow and the background state (hereafter the MJO component) and the QBW time-scale component are also obtained with a low-pass filter with a 20-day cutoff period and a bandpass filter with a 10–20-day period, respectively. In this study, wind fields are composited with respect to the TC center and the track-turning time.
So far, a generally accepted definition is not available for sudden TC track changes in the literature. In a composite study, Chan et al. (1980) examined the surrounding wind and temperature fields associated with track changes of TCs occurring in the West Indies during 1961–77. A minimum direction change of 20° within 12 h was used to define the left- and right-turning tracks. Although sudden north-turning cases are listed, no quantitative definition for the sudden track change was described in Carr and Elsberry (1995). In Wu et al. (2011a), the sharp track direction exceeded 60° over a 24-h period and the northwestward movement was replaced by the northeastward one. To identify sudden track changes from the JMA best-track data, we calculated all of the direction changes of TC movement in the WNP basin over the period 1970–2010 and found that the standard derivation of direction changes is about 28.5° and 24.2° for 12- and 6-h periods, respectively. In this study, the thresholds of track direction change are based on a compromise between the number of available samples and the degree of direction change. Considering the general westward or northwestward movement of TCs in the WNP basin, we take the threshold of direction change more than one standard derivation for the north-turning case. A sudden northward-turning track change is defined if a track direction change exceeds 40° (37°) during the 12-h (6 h) period. Despite a relative small change in direction, the west-turning track change represents the opposite track-turning type. In this study a west-turning track change is defined if a track direction change exceeds 25° (25°) during the 12-h (6 h) period. In both of the cases, we require that other track direction changes are relatively smaller than the thresholds over the 36-h period (i.e., 24 h before and 12 h after the turning time). If we face the direction of TC movement, the north-turning and west-turning track changes are similar to the right-turning and left-turning cases in Chan et al. (1980).
Based on the above definitions, track changes are identified west of 135°E in May–November during the 11-yr period (2000–10), including 15 sudden north-turning and 14 west-turning cases (Fig. 1). Note that the four cases of Saomai (2000), Maemi (2003), Sinlaku (2008), and Jangmi (2008) that were discussed in Wu et al. (2011a) are also included in the 15 sudden north-turning cases. Figure 2 shows the mean translation speed and direction of the TCs from 48 h before to 24 h after the turning time for the sudden north-turning cases and west-turning cases. The direction is measured clockwise from due west. On average the north-turning TCs take a general northwestward (~30°) track until 6 h prior to the sudden turning, and then a northward shift of about 90° (from ~30° to ~110°) occurs within 12 h. The translation speed decreases from ~5 to ~2 m s−1 at the turning time and then increases quickly with the northward shift in direction. The west-turning TCs also take a general northwestward track until 6 h prior to the turning time, following a sudden southward shift of ~40° (from ~40° to ~0°) within 12 h. While the speed changes are rather moderate, a remarkable feature in the west-turning cases is the persistent northward track shift during the 42 h prior to the turning time.
The 72-h segments of the identified (a) north-turning and (b) west-turning TC tracks during the period 2000–10. Here the track segments 48 h before and 24 h after turning are shown.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The 72-h segments of the identified (a) north-turning and (b) west-turning TC tracks during the period 2000–10. Here the track segments 48 h before and 24 h after turning are shown.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The 72-h segments of the identified (a) north-turning and (b) west-turning TC tracks during the period 2000–10. Here the track segments 48 h before and 24 h after turning are shown.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The 72-h mean (a),(b) translation speed and (c),(d) direction 48 h before and 24 h after the turning time for (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10, with the error bars indicating one standard deviation of the translation speed and direction in the corresponding bin.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The 72-h mean (a),(b) translation speed and (c),(d) direction 48 h before and 24 h after the turning time for (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10, with the error bars indicating one standard deviation of the translation speed and direction in the corresponding bin.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The 72-h mean (a),(b) translation speed and (c),(d) direction 48 h before and 24 h after the turning time for (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10, with the error bars indicating one standard deviation of the translation speed and direction in the corresponding bin.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
3. Analysis of the track forecast errors in NMC/CMA
Based on operational track forecasts from the NMC/CMA, the track forecast errors for seven sudden north-turning cases and nine west-turning cases during 2005–10 (earlier data are not available) are examined. For comparison, we first calculate the average errors for all of the track forecasts in NMC/CMA during 2005–10. With the respective sample sizes of 1637 and 1299, the average errors in distance are 112.6 km for 24-h forecasts and 188.5 km for 48-h forecasts. Figure 3 shows the track forecast errors for the seven north-turning cases and nine west-turning cases. The average 24- and 48-h forecast errors in distance are 145.6 and 317.3 km at the turning time for the sudden northward changes, increasing by 29.3% and 68.3%, respectively, compared to the average errors for all TCs.
Average track forecast errors for north-turning (black) and west-turning (gray) cases: (a) distance for 24-h forecasts, (b) distance for 48-h forecasts, (c) direction for 24-h forecasts, and (d) direction for 48-h forecasts. Solid lines in (a),(b) indicate the mean forecast errors in distance for all tropical cyclones during 2005–10.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Average track forecast errors for north-turning (black) and west-turning (gray) cases: (a) distance for 24-h forecasts, (b) distance for 48-h forecasts, (c) direction for 24-h forecasts, and (d) direction for 48-h forecasts. Solid lines in (a),(b) indicate the mean forecast errors in distance for all tropical cyclones during 2005–10.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Average track forecast errors for north-turning (black) and west-turning (gray) cases: (a) distance for 24-h forecasts, (b) distance for 48-h forecasts, (c) direction for 24-h forecasts, and (d) direction for 48-h forecasts. Solid lines in (a),(b) indicate the mean forecast errors in distance for all tropical cyclones during 2005–10.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As shown in Fig. 3, the distance error of 24-h (48 h) track forecasts is less than the average errors 12 h (30 h) prior to the northward turning. This is likely because the selected TCs kept moving westward or northwestward without substantial changes in direction before the northward turning. The forecasting error increases quickly and reaches a maximum of about 200 (400) km for the 24-h (48 h) forecasts 6 h (12 h) after the northward turning. The maximum 24- and 48-h direction errors are about 50° at 6 h. Typhoon Megi (2010) is an example of the failure in track direction forecasting. While all of the operational numerical models predicted that the typhoon would continue to move westward after passing the Philippines, its track suddenly turned northward with an angle more than 90° at 0000 UTC 20 October.
For the west-turning case, the 24- and 48-h forecast errors in distance are generally close to the average forecast errors with a relatively small error in direction. Note that the negative (positive) error in direction means an underprediction in the northward (westward) turning. This figure suggests that sudden northward track changes currently represent a challenge in TC forecasting in the WNP basin.
4. Relationship between sudden TC track changes and steering
It is well known that TC movement is primarily steered by the asymmetric flows in the vicinity of TC centers, which consist of the large-scale environmental flow and the ventilation flow that results from the interaction between TCs and their environment (Holland 1983; Carr and Williams 1989; Fiorino and Elsberry 1989; Wu and Wang 2000, 2001a,b). Here we show that the northward and westward track changes are also associated with changes in steering flows. We calculate the steering in this study as the mass-weighted mean wind averaged within a radius of 440 km between 850 and 300 hPa. Compared to the TC translation speed, Fig. 4 shows the calculated steering from 48 h before to 24 h after the turning time for the sudden north-turning and west-turning cases. As indicated in this figure, changes in the steering are consistent with those in TC translation. That is, we can understand the sudden track changes by examining mechanisms involved in the changes in the steering.
Comparisons of zonal Cx and meridional Cy components of the mean TC translation speed (gray) 48 h before and 24 h after the turning time (0 h) for (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10 with the corresponding steering (black) calculated between 850 and 300 hPa over a circle centered at the TC center with a radius of 440 km.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Comparisons of zonal Cx and meridional Cy components of the mean TC translation speed (gray) 48 h before and 24 h after the turning time (0 h) for (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10 with the corresponding steering (black) calculated between 850 and 300 hPa over a circle centered at the TC center with a radius of 440 km.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Comparisons of zonal Cx and meridional Cy components of the mean TC translation speed (gray) 48 h before and 24 h after the turning time (0 h) for (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10 with the corresponding steering (black) calculated between 850 and 300 hPa over a circle centered at the TC center with a radius of 440 km.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Moreover, the sudden track changes are reflected in the unfiltered low-level wind field associated with the TCs. Figures 5–7 show the 18-h evolution of the unfiltered 700-hPa winds prior to the track turning time. At −18 h (Fig. 5), the enhanced winds all appear on the northeast side in both of the north-turning and west-turning cases, indicating the generally northwestward movement of the TCs before the sudden track changes. At −6 h (Fig. 6), although the enhanced winds are still located to the northeast of the TC center in the west-turning case, the area of enhanced winds in the north-turning case shifts clockwise to the east of the TC center. At the turning time (Fig. 7), the maximum winds in the north-turning case are still located to the east of the TC center, but close examination indicates that the area of enhanced winds shifts counterclockwise in the west-turning case, in agreement with the changes in TC translation shown in Figs. 4b,c. Figures 5–7 suggest that the sudden track changes are consistent with the evolution of asymmetric wind patterns at the low levels.
Unfiltered 700-hPa winds (vectors, m s−1) and speeds larger than 16 m s−1 (contours and shading) composited with respect to TC centers for (a) the north-turning case and (b) the west-turning case 18 h prior to the turning time, with contour intervals of 2.0 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Unfiltered 700-hPa winds (vectors, m s−1) and speeds larger than 16 m s−1 (contours and shading) composited with respect to TC centers for (a) the north-turning case and (b) the west-turning case 18 h prior to the turning time, with contour intervals of 2.0 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Unfiltered 700-hPa winds (vectors, m s−1) and speeds larger than 16 m s−1 (contours and shading) composited with respect to TC centers for (a) the north-turning case and (b) the west-turning case 18 h prior to the turning time, with contour intervals of 2.0 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As in Fig. 5, but for 6 h prior to the turning time.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As in Fig. 5, but for 6 h prior to the turning time.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As in Fig. 5, but for 6 h prior to the turning time.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As in Fig. 5, but for the turning time T.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As in Fig. 5, but for the turning time T.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
As in Fig. 5, but for the turning time T.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
5. Multitime-scale flows associated with sudden track changes
Wu et al. (2011a) demonstrated that the cyclonic circulation of TCs over the western North Pacific can be coupled with low-frequency circulation and suggested that the low-frequency cyclonic gyres can interact with TCs, leading to sudden northward changes in TC tracks. They suggested that understanding of the contributions of various time-scale flows can shed light on sudden TC track changes. For this reason, the wind fields at the turning time in Fig. 7 are first decomposed into the ISO and synoptic components (Figs. 8 and 9).
700-hPa filtered wind fields and wind speeds larger than 8 m s−1 (contours and shading) for sudden north-turning cases at the turning time: (a) ISO component and (b) synoptic component, with contour intervals of 2 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa filtered wind fields and wind speeds larger than 8 m s−1 (contours and shading) for sudden north-turning cases at the turning time: (a) ISO component and (b) synoptic component, with contour intervals of 2 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa filtered wind fields and wind speeds larger than 8 m s−1 (contours and shading) for sudden north-turning cases at the turning time: (a) ISO component and (b) synoptic component, with contour intervals of 2 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa filtered wind fields and wind speeds larger than 8 m s−1 (contours and shading) for west-turning cases at the turning time: (a) ISO component and (b) synoptic component, with contour intervals of 2 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa filtered wind fields and wind speeds larger than 8 m s−1 (contours and shading) for west-turning cases at the turning time: (a) ISO component and (b) synoptic component, with contour intervals of 2 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa filtered wind fields and wind speeds larger than 8 m s−1 (contours and shading) for west-turning cases at the turning time: (a) ISO component and (b) synoptic component, with contour intervals of 2 m s−1. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
While the ISO wind fields in both of the north-turning and west-turning cases are associated with a large cyclonic gyre (Figs. 8a and 9a), the synoptic-scale wind fields show a remarkable difference in the spatial distribution of the wind speed. In the north-turning case (Fig. 8b), the strongest winds occur about 200 km east of the TC center, presenting a wavenumber-1 pattern with respect to the TC center in the spatial distribution of the synoptic wind speed. In the west-turning case (Fig. 9b), two wind maxima about 200 km away from the TC center can be seen on the southeast and northwest sides, respectively, indicating a wavenumber-2 pattern with respect to the TC center in the spatial distribution of the synoptic wind speed. The features in the ISO and synoptic-scale winds suggest their relative importance in the north-turning and west-turning cases during the sudden track changes. To demonstrate this, Fig. 10 shows the total, ISO, and synoptic-scale steering components for the north-turning and west-turning cases. In the north-turning case, the TC translation is dominated by the synoptic-scale steering since 6 h prior to the turning time because the wavenumber-1 pattern with respect to the TC center is dominant in the spatial distribution of the synoptic wind speed. In the west-turning case, on the other hand, the TC is mainly steered by the ISO flow since 18 h prior to the westward turning due to the dominance of the wavenumber-2 pattern with respect to the TC center in the spatial distribution of the wind speed. The steering effect of the enhanced southerly winds is nearly canceled by the northerly winds, leading to the dominance of the ISO steering. This may be one of the reasons for the small forecasting errors in the west-turning case.
The total (solid), ISO (short dashed), and synoptic-scale (long dashed) steering components in (a),(b) zonal and (c),(d) meridional directions 48 h before and 24 h after the turning time, composited from (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The total (solid), ISO (short dashed), and synoptic-scale (long dashed) steering components in (a),(b) zonal and (c),(d) meridional directions 48 h before and 24 h after the turning time, composited from (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The total (solid), ISO (short dashed), and synoptic-scale (long dashed) steering components in (a),(b) zonal and (c),(d) meridional directions 48 h before and 24 h after the turning time, composited from (a),(c) 15 sudden north-turning cases and (b),(d) 14 west-turning cases during 2000–10.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
The ISO gyre in the north-turning (west turning) case is elongated in the southwest-northeast (west–east) direction with the enhanced winds to the east (northeast) of the TC center. The relative locations of the enhanced ISO winds are linked to the subtropical high (Fig. 11). In the north-turning cases (Fig. 11a), with a weak trough to its north, the TC is located to the west of the subtropical high, thus the enhanced pressure gradient and winds appear to the east of the TC center. In the west-turning case (Fig. 11b), however, the subtropical high extends all the way westward to the north of the TC, resulting in enhanced winds in the northeast side of the ISO gyre.
Unfiltered 500-hPa geopotential height and wind vectors composited with respect to TC centers for (a) the north-turning case and (b) the west-turning case at the turning time. Contour intervals are 10 gpm and contours less than 5830 are not shown.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Unfiltered 500-hPa geopotential height and wind vectors composited with respect to TC centers for (a) the north-turning case and (b) the west-turning case at the turning time. Contour intervals are 10 gpm and contours less than 5830 are not shown.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Unfiltered 500-hPa geopotential height and wind vectors composited with respect to TC centers for (a) the north-turning case and (b) the west-turning case at the turning time. Contour intervals are 10 gpm and contours less than 5830 are not shown.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Lander (1994) suggested that two modes associated with TC formation in a monsoon gyre are 1) small (midget) TCs that form in the eastern periphery of a monsoon gyre, and 2) a giant TC that develops from the gyre itself. Examination of TC centers relative to the center of the ISO gyre at the turning time indicates that the TC centers of sudden north-turning cases are nearly collocated with the ISO gyre while the TC centers of west-turning cases are located to the east and northeast of the ISO gyre (figure not shown). It is implied that the track-turning types are related to the TC size. Carr and Elsberry (1995) suggested that the barotropic energy dispersion is sensitive to the monsoon gyre size. Based on baroclinic model simulations, Ge et al. (2008) found the Rossby wave energy dispersion is also associated with the TC intensity. However, since TC size data are not included in the current best-track datasets, it is hard to examine its influence on the track-turning types. The difference in the mean TC intensity between the north-turning cases and the west-turning cases are examined based on the maximum sustained wind data in the JMA best-track dataset. It is found that there is no statistically significant difference in the mean TC intensity between the two categories.
Figures 12 further shows the decomposed ISO wind fields. For the two specific types of sudden track changes, the composited TC circulation is embedded in a monsoon gyre of about 2500 km in diameter on the QBW scale (Lander 1994; Carr and Elsberry 1995). While the composited MJO flow pattern is mostly like a large cyclonic gyre in the north-turning cases and a monsoon trough in the west-turning cases, close examination indicates that a MJO cyclonic gyre can be identified in 12 north-turning cases and 8 west-turning cases. In the north-turning cases, the other three TCs are located at the bifurcation point of the MJO flows, as shown in Wu et al. (2011a). In the west-turning case, the other six TCs are associated with a monsoon trough on the MJO scale. In the north-turning case, an anticyclonic circulation can be clearly seen to the southeast of the QBW-scale gyre in Fig. 12b, suggesting the influence of the Rossby wave energy dispersion associated with the QBW-scale gyre. Carr and Elsberry (1995) argued that the Rossby wave energy dispersion of a cyclonic vortex is associated with the quantity 
700-hPa ISO wind fields for sudden north-turning cases and west-turning cases at the turning time: (a),(c) MJO time-scale components and (b),(d) QBW time scale. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa ISO wind fields for sudden north-turning cases and west-turning cases at the turning time: (a),(c) MJO time-scale components and (b),(d) QBW time scale. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa ISO wind fields for sudden north-turning cases and west-turning cases at the turning time: (a),(c) MJO time-scale components and (b),(d) QBW time scale. The coordinates are the latitudes and longitudes relative to the tropical cyclone center.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
6. Interactions between low-frequency and synoptic-scale circulations
Figure 13 shows the total 700-hPa synoptic-scale vorticity tendency for the north-turning case and the associated dominant terms calculated from the vorticity tendency equation in the framework moving with TCs. To focus on the peripheral ridging and troughing, the tendency is suppressed within a radius of 220 km due to uncertainty in the data associated with the TC inner circulation. For the north-turning case, the major ridging occurs to the southeast of the TC center, extending to about 600 km from the center (Fig. 13a). Although the resulting anticyclone is not clear in the synoptic wind field, the impact of the peripheral ridging can be inferred from the enhanced southwesterly winds 200–300 km away from the center because of the suppressing effect of the southeast ridging (Carr and Elsberry 1995). It can also be understood in terms of the resulting anomalous anticyclonic circulation, which leads to superposing additional southwesterly winds on the TC circulation on the southeast side. The ridging on the southeast side shown in Fig. 13a results mainly from the low-frequency (ISO) vorticity advection by synoptic flows, whereas the synoptic vorticity advection by low-frequency (ISO) flows tends to offset the ridging. Compared to the nonlinear interaction between the synoptic and low-frequency flows (Figs. 13b,c), the wavenumber-2 pattern of the stretching term suggests that it has little influence on the northward turning.
700-hPa synoptic-scale winds (m s−1) and vorticity tendency (10−10 s−2) calculated in the framework moving with TCs for north-turning cases at the turning time: (a) total, (b) ISO vorticity advection by synoptic flows, (c) synoptic vorticity advection by ISO flows, and (d) the stretching term.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa synoptic-scale winds (m s−1) and vorticity tendency (10−10 s−2) calculated in the framework moving with TCs for north-turning cases at the turning time: (a) total, (b) ISO vorticity advection by synoptic flows, (c) synoptic vorticity advection by ISO flows, and (d) the stretching term.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa synoptic-scale winds (m s−1) and vorticity tendency (10−10 s−2) calculated in the framework moving with TCs for north-turning cases at the turning time: (a) total, (b) ISO vorticity advection by synoptic flows, (c) synoptic vorticity advection by ISO flows, and (d) the stretching term.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
For the west-turning case (Fig. 14a), in addition to the primary ridging to the southeast of the TC, remarkable negative vorticity tendency can be also found to the northwest of the center. The enhanced southwesterly and northeasterly winds on the synoptic time scale in Fig. 9b are consistent with the ridging on the southeast and northwest sides of the TC. The dominant terms in the vorticity tendency are the low-frequency vorticity advection by synoptic flows (Fig. 14b) and synoptic vorticity advection by low-frequency flows (Fig. 14c), also indicating the nonlinear interaction between the ISO and synoptic-scale flows. While the southeast ridging in the north-turning cases results from both of the two terms, the northwest ridging shown in Fig. 14a is due to the ISO vorticity advection by synoptic flows, which is closely associated with the westward extension of the subtropical high in the west-turning case (Fig. 11b). The ISO flows, which are enhanced on the northeast side of the TC (Fig. 9a), lead to negative vorticity tendencies on the east side and positive vorticity tendencies on the northwest side through the low-frequency vorticity advection by synoptic flows (Fig. 14c). The latter partly offset the effect of the ISO vorticity advection by synoptic flows (Fig. 14b).
700-hPa synoptic-scale winds (m s−1) and vorticity tendency (10−10 s−2) calculated in the framework moving with TCs for west-turning cases at the turning time: (a) total, (b) ISO vorticity advection by synoptic flows, and (c) synoptic vorticity advection by ISO flows.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa synoptic-scale winds (m s−1) and vorticity tendency (10−10 s−2) calculated in the framework moving with TCs for west-turning cases at the turning time: (a) total, (b) ISO vorticity advection by synoptic flows, and (c) synoptic vorticity advection by ISO flows.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa synoptic-scale winds (m s−1) and vorticity tendency (10−10 s−2) calculated in the framework moving with TCs for west-turning cases at the turning time: (a) total, (b) ISO vorticity advection by synoptic flows, and (c) synoptic vorticity advection by ISO flows.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
Figure 15 shows the synoptic-scale vorticity tendency due to the low-frequency vorticity advection by synoptic flows for the north-turning case. At −18 h prior to the northward turning (Fig. 15a), when the TC is located to the south of the subtropical high, the vorticity tendency shows a wavenumber-2 pattern. This pattern is very similar to that in the west-turning case (Fig. 14b). The north ridging weakens at −12 h (Fig. 15b) and disappears after −6 h when the subtropical high lies to the east of the TC. At −6 h (Fig. 15c), the vorticity tendency pattern is dominated by a wavenumber-1 component. In other words, as shown in Fig. 9, as the wavenumber-1 component of the low-frequency vorticity advection by synoptic flows becomes dominant, the TC is primarily steered by the synoptic-scale flows.
700-hPa synoptic-scale vorticity tendency (10−10 s−2) due to ISO vorticity advection by synoptic flows for the north-turning cases: (a) −18, (b) −12, (c) −6, and (d) 0 h.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa synoptic-scale vorticity tendency (10−10 s−2) due to ISO vorticity advection by synoptic flows for the north-turning cases: (a) −18, (b) −12, (c) −6, and (d) 0 h.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
700-hPa synoptic-scale vorticity tendency (10−10 s−2) due to ISO vorticity advection by synoptic flows for the north-turning cases: (a) −18, (b) −12, (c) −6, and (d) 0 h.
Citation: Monthly Weather Review 141, 8; 10.1175/MWR-D-12-00224.1
7. Summary
Over the WNP and South China Sea, TCs are usually embedded in multitime-scale monsoon circulation (Ko and Hsu 2006, 2009) and TC tracks alternate between clusters of straight and recurving paths with an intraseasonal time scale (Harr and Elsberry 1991, 1995; Chen et al. 2009). Occasionally TCs experience sudden track changes (Carr and Elsberry 1995; Wu et al. 2011a). In this study, we identified 29 TCs that experienced track changes during the period 2000–10, including 15 sudden north-turning and 14 west-turning cases. Based on the operational track forecasts during the period 2005–10 from the NMC/CMA, with underprediction in the turning angle, it is found that the average 24- and 48-h distance errors at the turning time for the sudden north-turning case increase by 29.3% and 68.3%, respectively, compared to the average forecast errors for all TCs. It is suggested that sudden northward track changes currently represent a challenge in TC forecasting.
A composite analysis was conducted with the 15 sudden north-turning and 14 west-turning cases to reveal the influences of low-frequency flows on sudden track changes. For the two specific types of track changes, it is found that the composited TC circulation is embedded in a monsoon gyre of about 2500 km in diameter on the QBW scale, which is also embedded in a larger-scale cyclonic gyre or monsoon trough on the MJO scale. The TC motion during the turning period is generally consistent with changes in steering, which is a combination of the low-frequency (ISO) and synoptic-scale components in this study. During the track-turning period, it is found that the sudden northward turning is dominated by the synoptic-scale steering, whereas the west-turning case is controlled mainly by the low-frequency flow. This may be why sudden north-turning cases have larger track forecasting errors than west-turning cases.
Track changes are a result of interactions between the low-frequency and synoptic-scale flows (including TC circulation), which lead to asymmetric synoptic flow patterns that are significantly different between the north-turning and west-turning cases. In the north-turning case, winds are enhanced on the southeast side of TCs due to strong ridging associated with the low-frequency vorticity advection by synoptic flows, while the synoptic vorticity advection by low-frequency flows tends to offset the ridging effect. The southwesterly winds are enhanced 200–300 km away from the TC center because of the suppressing effect of the resulting ridging (Carr and Elsberry 1995). In the west-turning case, the westward-extended subtropical high leads to ridging on the northwest side of the TC through the low-frequency vorticity advection by synoptic flows, and thus enhances the synoptic-scale northeasterly winds. The enhanced winds on the west side can largely offset the steering of enhanced southwesterly winds. Thus, the west-turning case is controlled mainly by the low-frequency flow.
Acknowledgments
This research was jointly supported by the typhoon research project (2009CB421503) of the National Basic Research Program of China, the social commonweal research program of the Ministry of Science and Technology of China (GYHY200806009), the National Natural Science Foundation of China (NSFC Grant 41275093), and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).
REFERENCES
Carr, L. E., and R. T. Williams, 1989: Barotropic vortex stability to perturbations from axisymmetry. J. Atmos. Sci., 46, 3177–3191.
Carr, L. E., and R. L. Elsberry, 1995: Monsoonal interactions leading to sudden tropical cyclone track changes. Mon. Wea. Rev., 123, 265–290.
Chan, J. C. L., and R. T. Williams, 1987: Analytical and numerical studies of the beta-effect in tropical cyclone motion. Part I: Zero mean flow. J. Atmos. Sci., 44, 1257–1265.
Chan, J. C. L., W. M. Gray, and S. Q. Kidder, 1980: Forecasting tropical cyclone turning motion from surrounding wind and temperature fields. Mon. Wea. Rev., 108, 778–792.
Chan, J. C. L., W. X. Ai, and J. J. Xu, 2002: Mechanisms responsible for the maintenance of the 1998 South China Sea summer monsoon. J. Meteor. Soc. Japan, 80, 1103–1113.
Chang, C.-P., J. M. Chen, P. A. Harr, and L. E. Carr, 1996: Northwestward-propagating wave patterns over the tropical western North Pacific during summer. Mon. Wea. Rev., 124, 2245–2266.
Chen, T.-C., and J.-M. Chen, 1993: The 10–20-day mode of the 1979 Indian monsoon: Its relation with the time variation of monsoon rainfall. Mon. Wea. Rev., 121, 2465–2482.
Chen, T.-C., M.-C. Yen, and S.-P. Weng, 2000: Interaction between the summer monsoons in East Asia and the South China Sea: Intraseasonal monsoon modes. J. Atmos. Sci., 57, 1373–1392.
Chen, T.-C., S.-Y. Wang, M.-C. Yen, and A. J. Clark, 2009: Impact of the intraseasonal variability of the western North Pacific large-scale circulation on tropical cyclone tracks. Wea. Forecasting, 24, 646–666.
Duchon, C. E., 1979: Lanczos filtering in one and two dimensions. J. Appl. Meteor., 18, 1016–1022.
Fiorino, M., and R. L. Elsberry, 1989: Some aspects of vortex structure related to tropical cyclone motion. J. Atmos. Sci., 46, 975–990.
Ge, X., T. Li, Y. Wang, and M. Peng, 2008: Tropical cyclone energy dispersion in a three-dimensional primitive equation model: Upper-tropospheric influence. J. Atmos. Sci., 65, 2272–2289.
Goswami, B. N., R. S. Ajayamohan, P. K. Xavier, and D. Sengupta, 2003: Clustering of synoptic activity by Indian summer monsoon intraseasonal oscillations. Geophys. Res. Lett., 30, 1431, doi:10.1029/2002GL016734.
Harr, P. A., and R. L. Elsberry, 1991: Tropical cyclone track characteristics as a function of large-scale circulation anomalies. Mon. Wea. Rev., 119, 1448–1468.
Harr, P. A., and R. L. Elsberry, 1995: Large-scale circulation variability over the tropical western North Pacific. Part I: Spatial patterns and tropical cyclone characteristics. Mon. Wea. Rev., 123, 1225–1246.
Holland, G. J., 1983: Tropical cyclone motion: Environmental interaction plus a beta effect. J. Atmos. Sci., 40, 328–342.
Holland, G. J., and R. T. Merrill, 1984: On the dynamics of tropical cyclone structural-changes. Quart. J. Roy. Meteor. Soc., 110, 723–745.
Kiladis, G. N., and M. Wheeler, 1995: Horizontal and vertical structure of observed tropospheric equatorial Rossby waves. J. Geophys. Res., 100 (D11), 22 981–22 997.
Ko, K.-C., and H.-H. Hsu, 2006: Sub-monthly circulation features associated with tropical cyclone tracks over the East Asian monsoon area during July–August season. J. Meteor. Soc. Japan, 84, 871–889.
Ko, K.-C., and H.-H. Hsu, 2009: ISO modulation on the submonthly wave pattern and the recurving tropical cyclones in the tropical western North Pacific. J. Climate, 22, 582–599.
Krishnamurti, T. N., and P. Ardanuy, 1980: The 10 to 20-day westward propagating mode and “breaks in the monsoons.” Tellus, 32, 15–26.
Lander, M. A., 1994: Description of a monsoon gyre and its effects on the tropical cyclones in the western North Pacific during August 1991. Wea. Forecasting, 9, 640–654.
Lau, K.-H., and N.-C. Lau, 1990: Observed structure and propagation characteristics of tropical summertime synoptic scale disturbances. Mon. Wea. Rev., 118, 1888–1913.
Liang, J., L. Wu, X. Ge, and C.-C. Wu, 2011: Monsoonal influence on typhoon Morakot (2009). Part II: Numerical study. J. Atmos. Sci., 68, 2222–2235.
Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708.
Madden, R. A., and P. R. Julian, 1972: Description of global-scale circulation cells in tropics with a 40–50 day period. J. Atmos. Sci., 29, 1109–1123.
Madden, R. A., and P. R. Julian, 1994: Observations of the 40–50-day tropical oscillation—A review. Mon. Wea. Rev., 122, 814–837.
Merrill, R. T., 1984: A comparison of large and small tropical cyclones. Mon. Wea. Rev., 112, 1408–1418.
Murakami, M., 1976: Cloudiness fluctuations during summer monsoon. J. Meteor. Soc. Japan, 54, 175–181.
Murakami, M., and M. Frydrych, 1974: On the preferred period of upper wind fluctuations during the summer monsoon. J. Atmos. Sci., 31, 1549–1555.
Wu, L., and B. Wang, 2000: A potential vorticity tendency diagnostic approach for tropical cyclone motion. Mon. Wea. Rev., 128, 1899–1911.
Wu, L., and B. Wang, 2001a: Movement and vertical coupling of adiabatic baroclinic tropical cyclones. J. Atmos. Sci., 58, 1801–1814.
Wu, L., and B. Wang, 2001b: Effects of convective heating on movement and vertical coupling of tropical cyclones: A numerical study. J. Atmos. Sci., 58, 3639–3649.
Wu, L., H. Zong, and J. Liang, 2011a: Observational analysis of sudden tropical cyclone track changes in the vicinity of the East China Sea. J. Atmos. Sci., 68, 3012–3031.
Wu, L., J. Liang, and C.-C. Wu, 2011b: Monsoonal influence on typhoon Morakot (2009). Part I: Observational analysis. J. Atmos. Sci., 68, 2208–2221.
Yasunari, T., 1981: Structure of an Indian summer monsoon system with around 40-day period. J. Meteor. Soc. Japan, 59, 336–354.
